Refine
Year of publication
Document Type
- Doctoral Thesis (18)
Has Fulltext
- yes (18)
Is part of the Bibliography
- no (18)
Institute
- Biowissenschaften (11)
- Medizin (7)
Stem cells are often referred to as potential candidates for the treatment of different pathologies. Their ability to differentiate into various tissue specific cell types offers the possibility to engineer cell systems or organs for replacement. One of the main questions in stem cell biology is how stemness properties are regulated and to what extend this regulation is intrinsic or conveyed by the direct microenvironment (‘niche’). In order to elucidate such regulatory processes, it is informative to analyze processes or molecules that are shared between different stem cell populations.
One such molecule that is expressed on a wide range of different embryonic and adult as well as tumor stem cells is the ABC transporter Abcg2. ABC transporters in general are transmembrane proteins that actively extrude endo- and exotoxins as well as xenobiotics, thereby protecting cells and organs. Additionally, ABC transporters are responsible for drug resistance in many cancers. A well-described characteristic of stem cells expressing Abcg2 is the formation of the ‘side population’ (SP) phenotype: An active Abcg2 transporter mediates the efflux of a particular fluorescent dye that is taken up by all cells, thus leading to a less brightly stained population. This phenomenon is widely used to characterize and isolate the most primitive stem cell subpopulation from embryonic and adult tissues, including tumors. Besides its role as toxin transporter little is known about the function of Abcg2 in stem cells. This is mainly due to the fact that its physiological substrate in stem cells remains unknown. The identification of such substrates is therefore of high interest because it would directly link the activity of ABC transporters to regulatory mechanisms in stem cell biology.
In the present study we wanted to test the hypothesis that the sphingolipid ceramide is a physiological substrate of the ABC transporter Abcg2. Sphingolipids are potent second messengers and are known to have regulatory functions in stem cells. In particular, the sphingolipid ceramide is described as a mediator of controlled cell death and inducer of differentiation. It is suggested that stem cells need to keep their intracellular ceramide content at low levels in order to prevent apoptosis or differentiation. We propose that Abcg2 and ceramide interact and that this interaction leads to changes in the absolute or relative amounts of ceramide. This in turn influences basic stem cell functions such as self renewal and differentiation.
We show that Abcg2 prevents cells from accumulating fluorescence labeled ceramide. Furthermore, exogenously applied ceramides inhibit the transport activity of Abcg2, measured by a decrease of the side population phenotype. This inhibitory effect is consistent with a competitive inhibition mechanism. Additionally, we show that active Abcg2 can increase the ceramide concentration in cell culture supernatant. Finally we demonstrate that Abcg2 protects from ceramide induced cytotoxicity in human cell lines. In summary, these in vitro results strongly suggest that Abcg2 has the ability to regulate ceramide levels.
Murine hematopoietic stem cells (HSCs) are the best characterized adult stem cell system so far. By using 7-colour fluorescence-activated cell sorting (FACS) we established the purification of the most primitive HSCs, reflected by their high engraftment capability when transplanted to lethally irradiated mice. By using this sorted cell populations it was in addition possible to establish a system to reproducibly manipulate HSCs ex vivo. This experimental system will serve in further elucidating the physiological consequences of Abcg2 mediated changes in ceramide levels on stem cells in vivo.
Taken together, this study shows that Abcg2 has the ability to regulate ceramide levels in cells. This in turn can lead to cellular protection from ceramide induced apoptosis. Additionally, the experimental techniques to further analyze the role of Abcg2 and ceramide in the most primitive hematopoietic stem cells were successfully established, enabling more detailed analysis in the future.
Glioblastoma is the most common and most aggressive type of brain tumor in adults. In contrast to epithelial cancers, glioblastomas do not metastasize. While the major treatment challenge in epithelial cancers is not the primary tumor but metastasis, glioblastoma patients die of the primary tumor.
However, there is a common theme which underlies the malignant properties of progressed epithelial cancers and glioblastoma: invasion from the primary tumor into the surrounding tissue. In the case of epithelial cancers this is the first and necessary step to metastasis, whereas invasion leads inevitably to tumor recurrence after resection in the case of glioblastoma, causing it to be incurable.
A cellular program which has been described in detail to promote the invasive phenotype in epithelial tumors, is the epithelial-mesenchymal-transition (EMT). Differentiated neural cells are not epithelial, thus, strictly speaking, EMT does not occur in glioblastoma. However, the traits acquired in the process of EMT, especially invasiveness and stemness, are highly relevant to glioblastoma. One of the key transcription factors known to induce EMT in epithelial cancers is ZEB1, which has been described only marginally in the central nervous system so far. Here, I investigate the expression and function of ZEB1 in glioblastoma and during human fetal neural development.
ZEB1 mRNA was significantly upregulated in all histological types of glioma, including glioblastoma, when compared to normal brain. There was no correlation between ZEB1 mRNA levels and tumor grade. Immunohistochemical staining of glioma samples demonstrated that ZEB1 was highly expressed in the great majority of tumor cells. In the developing human brain, intense staining for ZEB1 could be observed in the ventricular and subventricular zone, where stem- and progenitor cells reside. ZEB1 positive cells included cells stained with stem- and progenitor markers like PAX6, GFAP and Nestin. In contrast, ZEB1 was never found in early neuronal cells as identified by TUBB3 staining.
To gain insight into ZEB1 function I generated a human fetal neural stem cell line and a glioblastoma cell line with ZEB1 knockdown, which were compared with their respective control cell lines. First, I found that ZEB1 does not regulate the micro RNA 200 family in either cell line, which has been described as an essential ZEB1 target in epithelial cancers. Second, regulated target genes were identified with a genome wide microarray. The third approach was to directly identify genomic binding sites of ZEB1 by chromatin immunoprecipitation sequencing (ChIP-seq). All three approaches showed that the ZEB1 transcriptional program is surprisingly similar in the neural stem cell line and the glioblastoma cell line. In contrast, it bears only little resemblance to the program described in epithelial cancers.
The most interesting, previously unrecognized ZEB1 target gene identified in this study is integrin b1. It was regulated after ZEB1 knockdown detected by microarray analysis, and has a ZEB1 binding site in its promoter region detected by ChIP-seq. Finally, I addressed the question whether ZEB1 influences tumor growth and invasiveness in a glioblastoma model. After intracranial xenotransplantation in mice, ZEB1 knockdown glioblastoma cells formed significantly smaller and less invasive tumors than control glioblastoma cells.
This study demonstrates that ZEB1 is widely expressed in glioma and relevant for glioblastoma growth and invasion. In contrast to what is known about ZEB1 function in epithelial cancers, ZEB1 is not associated with glioma progression, but instead seems to be an early and necessary event in tumorigenesis. Also with regard to ZEB1 target genes, ZEB1 functions differently in glioblastoma than in epithelial cancers. The two most important ZEB1 targets in epithelial cancers are E-cadherin and the miR-200 family members. Both are not relevant to ZEB1 function in glioblastoma. Interestingly, while the ZEB1 transcriptional program is different from the one described in epithelial cancers, it is highly similar in glioblastoma cells and fetal neural stem cells. This suggests that an embryonic pathway restricted to stem- and progenitor cells during development is reactivated in glioblastoma.
Previously known ZEB1 target genes were tissue specific and therefore seemed unlikely to mediate ZEB1 function in the central nervous system. However, the newly identified ZEB1 target gene integrin b1 is well known to play pivotal roles in both glioblastoma tumorigenesis and invasion as well as in neural stem cells. Additionally, integrin b1 is widely expressed and seems a likely ZEB1 target in other organs than the brain.
Taken together, I demonstrate that ZEB1 is a new regulator of glioblastoma growth and invasion. The transcriptional program of ZEB1 differs from the one in epithelial cancers but is strikingly similar to the one in neural stem cells. The newly identified ZEB1 target gene integrin b1 is likely to mediate crucial ZEB1 functios. Thus, this study identifies ZEB1 as a yet unrecognized player in glioblastoma and neural development. Furthermore, it sets the stage for more research which will help to deepen our understanding of ZEB1 function in the central nervous system and beyond.
Einleitung In ihrer klinischen Symptomatik lassen sich der ischämische Schlaganfall (IS) und die intrazerebrale Blutung (ICH) nicht sicher unterscheiden. Hinsichtlich der Akuttherapie, die eine möglichst schnelle Wiederherstellung der zerebralen Sauerstoffversorgung („time is brain“) zum Ziel hat, ist diese Differenzierung jedoch essentiell. Das Ziel der vorliegenden Arbeit ist der Nachweis serologischer Biomarker in der Frühphase des Schlaganfalls zur Differenzierung zwischen IS und ICH. Hypothesen-gestützt wurden aufgrund pathophysiologischer Überlegungen hierfür die ZNS-spezifischen Proteine GFAP, UCH-L1, pNf-H, MBP und Tau untersucht. In einem hypothesenfreien Ansatz wurden Serumproben von Schlaganfallpatienten massenspektrometrisch analysiert.
Material und Methoden Die Patientenrekrutierung für die vorliegende Arbeit erfolgte im Rahmen der prospektiven, multizentrischen BE FAST II-Studie zur Evaluation von GFAP als Biomarker der akuten ICH. Von Mai 2012 bis April 2014 wurden Serumproben von Patienten mit akuter Schlaganfallsymptomatik in der Klinik für Neurologie der Goethe-Universität Frankfurt am Main gesammelt. Mittels kommerziell verfügbaren ELISA-Kits wurden die Serumkonzentrationen der Proteine UCH-L1, pNf-H, MBP und Tau bestimmt. Die Analyse der GFAP-Serumkonzentration erfolgte durch Roche Diagnostics mit Hilfe des Elecsys®-Systems, einem Elektrochemiluminiszenz-Immunoassay. Für die Massenspektrometrie wurden Serumproben aus der BE FAST-I-Studie, die von Ende des Jahres 2010 bis Anfang des Jahres 2011 asserviert wurden, eingesetzt. Die massenspektrometrischen Untersuchungen wurden in der Abteilung „Funktionelle Proteomics“ im Fachbereich Medizin der Goethe-Universität Frankfurt am Main durchgeführt.
Ergebnisse Tau und MBP ließen sich in den meisten Serumproben nicht nachweisen. In der pNf-H-Messung lag die Konzentration bei 27 von 35 Patienten oberhalb der Nachweisgrenze, wobei sich kein signifikanter Unterschied zwischen IS und ICH ergab (p = 0,69). UCH-L1 ließ sich bei 28 von 29 Patienten nachweisen. In der IS-Gruppe war eine signifikant (p = 0,005) höhere UCH-L1-Konzentration nachweisbar (Median 5,71 ng/ml) als in der ICH-Gruppe (Median 2,37 ng/ml). GFAP war bei allen 45 Patienten nachweisbar mit signifikant (p < 0,00005) höherer Konzentration in der ICH-Gruppe (Median 2,87 ng/ml) verglichen mit der IS-Gruppe (Median 0,01 ng/ml). Zudem fand sich eine positive Korrelation der UCH-L1-Werte in der IS-Gruppe mit dem Patientenalter (r = 0,62, p = 0,01), sowie eine positive Korrelation der GFAP-Werte in der ICH-Gruppe mit dem Patientenalter (r = 0,54, p = 0,03), dem NIHSS-Wert (r = 0,69, p = 0,04) und mit dem ICB-Volumen (r = 0,60, p = 0,01). In der massenspektrometrischen Analyse ließ sich eine Top Liste aus 22 Proteinen erstellen, die jeweils signifikante Unterschiede zwischen IS und ICH aufweisen.
Diskussion Die Rolle des Ubiquitin-Proteasom-System (UPS) und insbesondere von UCH-L1 beim IS ist bislang noch nicht abschließend geklärt. Nach einer zerebralen Ischämie ist jedoch eine Upregulation von UCH-L1 beschrieben, u.a. durch eine verstärkte UPS-Aktivität durch Aggregate fehlgefalteter Proteine. Daneben reagieren Neurone sensibler auf eine Hypoxie als Gliazellen mit einer dominierenden Freisetzung neuronaler Proteine wie UCH-L1. Bei ICH kommt es dagegen eher zu einer unspezifischen Destruktion des Hirngewebes mit vorwiegender glialer Schädigung und rascher Freisetzung glialer Proteine wie GFAP. Mit UCH-L1 und GFAP konnten zwei Proteine als erfolgsversprechende Kandidaten zur Differenzierung zwischen IS und ICH in der Frühphase identifiziert werden. Zur weiteren Validierung sind Untersuchungen an einer großen Population notwendig, die auch kleinere Infarkte und Hirnblutungen einschließt. Auch der Einfluss epidemiologischer und klinischer Faktoren wie z.B. dem Patientenalter muss weiter evaluiert werden.
Die mittels Massenspektrometrie erstellte Top Liste aus 22 Proteinen enthält vielversprechende Biomarker-Kandidaten, die signifikante Unterschiede zwischen IS und ICH aufweisen und ebenso an einem großen Patientenkollektiv weiter untersucht werden müssen.
Glioblastoma multiforme accounts for more than 80% of all malignant gliomas in adults and a minor fraction of new annual cases occurs in children. In the last decades, research shed light onto the molecular patterns underlying human malignancies which resulted in a better understanding of the disease and finally an improved long term survival for cancer patients. However, malignancies of the central nervous system and especially glioblastomas are still related to poor outcomes with median survivals of less than 6 months despite extensive surgery, chemotherapy and radiation. Hence, a better understanding of the molecular mechanism driving and sustaining cancerous mutations in glioblastomas is crucial for the development of targeted therapies. Apoptosis, a form of programmed cell death, is an important feature of eukaryotic cells and crucial for the maintenance of multicellular homeostasis. Because apoptosis is a highly complex and tightly regulated signaling pathway, resisting apoptotic stimuli and avoiding cell death is a hallmark of the cancerous transformation of cells. Hence, targeting molecular structures to reestablish apoptotic signaling in tumor cells is a promising approach for the treatment of malignancies. Smac mimetics are a group of small molecular protein inhibitors that structurally derive from an intracellular protein termed Smac and selectively block Inhibitor of apoptosis (IAP) proteins, which are often aberrantly expressed in cancer. Several studies confirmed the antitumoral effects of Smac mimetics in different human malignancies, including glioblastoma, and give rationales for the development of potent Smac mimetics and Smac mimetic-based combination protocols. This study investigates the antitumoral activity of the bivalent Smac mimetic BV6 in combination with Interferon α. Latter is a well characterized cytokine with an essential role in immunity, cell differentiation and apoptosis. This study further aims to address the molecular mechanisms underlying the antitumoral activity of the combination treatment by using well established molecular cell death assays, flow cytometry, western blot analysis, genetic approaches and selective pharmacological inhibition. Since different Smac mimetics and Smac mimetic-based combination therapies are currently under clinical evaluations, findings of this study may have broad implications for the application of Smac mimetics as clinical cancer therapeutics.
Beim ischämischen Schlaganfall finden weitreichende systemische immunmodulatorische Anpassungsvorgänge statt. Da Sphingosin-1-Phosphat (S1P)-Signalwege für die Immunzellrekrutierung von hoher Relevanz sind, war angesichts der bekannten immunologischen Veränderungen nach zerebraler Ischämie das Ziel dieser Dissertation die genauen Veränderungen dieses Signalweges zu charakterisieren.
Für diese Charakterisierung wurde ein transientes Fadenokklusionsmodell der A. cerebri media an der Maus verwendet. Die Sphingolipidkonzentrationen wurden drei oder 24 Stunden nach Okklusion in der Milz, im Plasma sowie im Hirngewebe gemessen. Parallel hierzu wurde die Immunzellrekrutierung in die von der Ischämie betroffenen Hemisphäre analysiert.
Zunächst konnte diese Dissertation zeigen, dass in der Akutphase des Schlaganfalls ein S1P-Konzentrationsgradient vorherrscht. Die Milz zeigt hier die niedrigsten Konzentrationen, gefolgt von Plasma und Gehirn. Darüber hinaus besteht auch in der betroffenen Hemisphäre ein S1P-Gradient mit hohen Konzentrationen im Infarktkern, jedoch verminderten Konzentrationen im Periinfarktkortex (PIC).
Zweitens führt eine fokale zerebrale Ischämie zu einer Infiltration von T- und B-Lymphozyten in die ischämische Hemisphäre. Im Gegensatz hierzu kommt es zu einer Schlaganfall-induzierten Lymphopenie im Blut. Hierzu passend konnte ich eine signifikante Abnahme des Gewichts und der B- und T-Lymphozyten der Milz 24 Stunden nach Ischämie nachweisen. Weitere von Immunzellen produzierte Zytokine (IL-6) sowie deren Transkriptionsfaktoren (SPI1, STAT3, FoxP3) zeigten in der Akutphase nach Ischämie ebenfalls eine deutliche Reduktion und wiesen auf die Rekrutierung peripherer Immunzellen (pIZ) aus dem sekundären lymphatischen Organ hin. Folgerichtig waren Leukozyten im Plasma sowohl drei als auch 24 Stunden nach Ischämie signifikant vermehrt, welche insbesondere neutrophilen Granulozyten entsprachen.
Basierend auf der nachgewiesenen Reduktion von T-Helferzellen sowie regulatorischer T-Zellen sowohl in der Milz als auch in der Zirkulation, wurde drittens die Hypothese einer zerebralen Rekrutierung dieser T-Zellpopulationen gemäß dem vorliegenden S1P-Gradienten untersucht. Dabei gelang die Darstellung einer signifikanten Infiltration von CD45+-Zellen in beide Hemisphären, welche insbesondere von T-Helferzellen geprägt war.
Viertens nimmt die S1P-Rezeptor (S1PR)-Expression auf Leukozyten eine bedeutende Stellung in der pIZ-Rekrutierung ein. In diesem Sinne konnte ich zeigen, dass nach zerebraler Ischämie S1P1 signifikant in der Milz vermindert exprimiert wurde. Dieses Ergebnis deutete auf einen Austritt S1P1+ Immunzellen aus der Milz dem etablierten S1P-Gradienten folgend hin. In der ischämischen Hemisphäre hingegen ließ sich ebenfalls eine Herunterregulation der exprimierten mRNA für S1P1 nachweisen, wohingegen S1P2 und S1P3 vermehrt transkribiert wurden. Dieses Ergebnis könnte Folge der mikroglialen Aktivierung sein, die bekanntermaßen mit einer Hochregulation von S1P2 und S1P3 einhergeht.
Abschließend habe ich die Rolle von weiteren Sphingolipiden, u.a. von Ceramiden, untersucht, die einen signifikanten Anstieg in der Milz 24 Stunden nach Ischämie zeigten. Im Gegensatz dazu konnte ich im Gehirn keine Unterschiede der untersuchten Ceramidspezies abgrenzen, sodass in dem hier verwendeten Modell eine Beteiligung an lokalen pathophysiologischen Vorgängen eher unwahrscheinlich erscheint.
Zusammenfassend beschreiben die in dieser Dissertation dargestellten Ergebnisse lokale und systemische Veränderungen des S1P-Signalwegs nach zerebraler Ischämie. Konkordante Veränderungen des Immunsystems deuten auf eine relevante Rolle veränderter S1P-Konzentrationen hin. Weitergehende, funktionelle Untersuchungen der hier beobachteten Ergebnisse müssen die potentielle therapeutische Relevanz für Patienten mit zerebraler Ischämie aufklären.
The canonical Wnt/β-catenin and the Shh pathway as well as the Notch signaling cascade
are key regulators in stem cell biology and are independently associated with the development
of cancer. Despite the knowledge of a balanced signaling for cellular maintenance, the
fundamental biochemical mechanisms of crosstalk are still poorly understood. This study
demonstrates that the outcome of interaction between Wnt and Shh is cell type specific. A
combined inhibitory mechanism of the Shh and Notch2/Jagged2 pathways on dominant
active β-catenin signaling in the adult tongue epithelium keeps Wnt/β-catenin signaling
restricted to physiological tolerable levels. In the opposite crosstalk the activation of
Wnt/β-catenin signaling in medulloblastoma (MB) of the Shh subtype, in turn inhibits the Hh
pathway.
The inhibitory mechanism of Shh and Notch2/Jagged2 on Wnt/β-catenin signaling is
independent of the degradation complex of β-catenin and takes place inside the nucleus.
Furthermore, the negative feedback on Wnt/β-catenin signaling by the Shh pathway relies
on transcriptional activity of Gli1/2A. Inhibition of Gli1/2A with the specific inhibitor GANT61
abrogated the negative impact of Shh on β-catenin signaling in vitro. Although the negative
feedback loop of Shh is still functional in human SCC25 cells, the inhibitory effect of
Notch2/Jagged2 is lost and contributes to the cancerogenic phenotype of these cells. In the
inverse situation, the activation of β−catenin signaling has a negative feedback on
constantly active Shh signaling and significantly inhibits the Hh pathway. This was shown in
Ptch+/- and Math1-Cre:SmoM2Fl/+ MB tumor spheres in vitro, in which inhibition of sphere
formation and growth was observed and Hh target gene transcription was down-regulated.
This demonstrates for the first time that the activation of canonical Wnt/β-catenin signaling
in primary MB cells with a Hh pathway over-activation has a negative effect on the growth of
these cells in vitro.
In summary the results show that crosstalk of Wnt/β-catenin and Shh signaling has context
specific outcome on pathway activity. Elucidation of the molecular interactions will improve
our understanding of Wnt and Hh associated tumors and contribute to the development of
new therapeutic strategies.
Blood vessel formation is a well orchestrated process where multiple components including different cells types, growth factors as well as extracellular matrix proteins act in synergistic and highly regulated manner to support the growth of new blood vessels. During embryonic development this process is marked as vasculogenesis and entails the differentiation of mesodermal cells into angioblasts and their subsequent fusion into a primitive vascular plexus. Angiogenesis, in contrast, describes the formation of new vessels from the pre-existing vasculature and it occurs in the embryo during remodeling of the primitive plexus into a mature vascular network. Furthermore, in the adult, angiogenic processes play a role in various physiological and pathological conditions. Angiogenesis is governed by a set of factors and molecular mechanisms whose identification has been a major focus of cardiovascular research for the past several decades. Most recently, Epidermal growth factor-like domain 7 (EGFL7) has been described as a novel molecular player in this context. This secreted protein is produced by endothelial cells and has been implicated in vessel development. Studies performed in zebrafish revealed an important role for EGFL7 in lumen formation during vasculogenesis although the underlying molecular mechanism has not been elucidated yet. In contrast, the investigation of EGFL7’s functions during angiogenic sprouting has faced several challenges and the role of EGFL7 in angiogenesis remained elusive. The purpose of this thesis was to identify the functions of EGFL7 during angiogenic mode of vessel formation in a systematic fashion using numerous in vitro as well as in vivo approaches.
Previously it has been suggested that EGFL7 might associate with the extracellular matrix from where it could exert its effects. Indeed, we could show that EGFL7 accumulates on the outer surface of endothelial cells in vivo by demonstrating its co-localization with collagen IV, a major constituent of the basal lamina. Furthermore, after its secretion to the extracellular matrix (ECM), EGFL7 seemed to interact with some components of the extracellular matrix including fibronectin and vitronectin, but not collagens and laminin.
A major group of receptors that mediate the interaction between the cells and the ECM are integrin receptors. Our co-immunoprecipitation studies revealed that EGFL7 associated with integrin αvβ3 which is highly expressed in endothelial cells and known to be important for vessel growth. Importantly, this EGFL7-αvβ3 integrin interaction was dependent on Arg-Gly-Asp (RGD) motif present within the second EGF-like domain of EGFL7 protein. Adhesion assays performed with human umbilical vein endothelial cells (HUVEC) revealed that EGFL7 promoted endothelial cell adhesion compared to BSA used as a negative control, however, adhesion seemed to be less efficient as compared to bona fide ECM proteins such as fibronectin and vitronectin. In addition, cultivation of endothelial cells on EGFL7 was characterized by the absence of mature focal adhesions and stress fibers, but was paralleled by increased phosphorylation of kinases typical for integrin activation signaling cascade such as FAK, Src and Akt. This led us to the hypothesis that EGFL7 creates an environment that supports a motile phenotype of endothelial cells by serving as a modulator of existing interactions between the cells and the surrounding matrix. Indeed, EGFL7 increased random migration of HUVEC on fibronectin in an αvβ3 integrin dependent manner as shown using a live cell imaging platform. Most importantly, this was paralleled by a decrease in endothelial cell adhesion to fibronectin which is consistent with previous reports on secreted proteins that support a medium strength of adhesion and such promote cellular migration. To assess the overall effect of EGFL7 on the process of blood formation several in vitro and in vivo approaches were employed. First, the addition of EGFL7 to Matrigel injected subcutaneously into mice significantly increased the invasion of endothelial cells into the plugs. Second, a spheroid-based sprouting assay in three-dimensional collagen matrix clearly demonstrated the ability of EGFL7 to support angiogenic sprouting in an integrin dependent manner. This is consistent with the observed effects of EGFL7 on endothelial cell migration. Third, using in vivo assays such as the chick chorioallantoic membrane (CAM) assay as well as a zebrafish model system we were able to validate the importance of the EGFL7-integrin interaction for the process of angiogenesis in vivo. Taken together, I identified some of the major cellular functions EGFL7 modulates during angiogenesis. In addition, with integrin αvβ3 I unraveled a novel interaction partner of EGFL7 that delivers a mechanistical explanation for EGFL7’s effects on blood vessel formation. Most importantly, data presented in this PhD thesis contribute substantially to the existing literature on EGFL7 unambiguously assigning a role for this protein in the process of angiogenesis.
The canonical Wnt pathway, also known as Wnt/β-‐catenin pathway, comprises a network of proteins which control diverse developmental and adult processes in all metazoan organisms. The binding of canonical Wnt ligands to a cell surface receptor complex, consisting of frizzled family members and low density lipoprotein receptor-‐ related protein 5 or 6 co‐receptors, triggers a signaling cascade which results in a β-catenin-‐mediated transcriptional activation of different target genes, implicated in cellular proliferation, apoptosis, migration and differentiation. A couple of years ago, several groups including us, iden2fied transient activation of the canonical Wnt-pathway in endothelial cells (ECs) of the developing central nervous system (CNS). In this context, Wnt/β-‐catenin signaling could be demonstrated to be crucial for brain angio genesis as well as for the establishment of the blood-brain barrier (BBB) phenotype in the newly formed vessels.
Gliomas, in particular the glioblastoma (GBM), belong to the group of highly vascularized solid tumors which gain their vascularization due to an angiogenic switch occurring during tumor progression. Interestingly, nuclear localized β-‐catenin could be exclusively detected in the activated endothelium of induced rat gliomas and of human GBM, suggesting a so far unknown and not further characterized involvement of the canonical Wnt pathway in pathological angiogenesis. In order to systematically decipher the precise role of endothelial Wnt/β-‐catenin signaling in tumor angiogenesis, I established
murine GL261 glioma cell lines overexpressing either Wnt1 or Dickkopf (Dkk) 1 in a doxycycline-‐dependent manner, an activator and potent inhibitor of Wnt/β-‐catenin signaling, respectively. In subcutaneous and intracranial transplantations, tumor-derived Wnt1 reduced, while Dkk1 increased GL261 tumor growth without affecting in vitro proliferation, cell cycle or cell death of the established cell lines. Nowadays, it is well accepted that solid tumors are dependent on vascular support allowing them to grow beyond a certain size. In my work I could show that tumor-‐derived Wnt1 targets the tumor vasculature by increasing endothelial Wnt/β-‐catenin signaling, which reduced tumor vessel density and resulted in a more quiescent tumor vasculature. Furthermore, Wnt1-‐expression mediated tight association of smooth muscle cells (SMCs) and pericytes to the tumor endothelium, a phenotype which is unusual for tumor vessels and a described hallmark of tumor vessel normalization. In contrast, inhibition of endothelial Wnt/β-‐catenin signaling by Dkk1 mediated an opposing effect, characterized by endothelial hyper-proliferation and a tumor vasculature with a rough basal lamina distribution and loosely anached mural cells, indicative of a strong angiogenic activity. The described vascular effects in Wnt1-expressing GL261 tumors could be verified by subcutaneous transplantations of a rat glioma cell line constitutively expressing Wnt1. Furthermore, an applied in vivo MatrigelTM plug assay uncovered the reduction in vessel density upon Wnt1 simulation to be tumor cell independent, suggesting an EC-‐autonomous effect. This hypothesis was confirmed by subcutaneous transplantations of parental GL261 cells into mice with genetically generated endothelial β-‐catenin gain-of-function (GOF). The derived GOF tumor from this experiment comprised a quiescent and normalized tumor vasculature and phenocopied the vascular effects observed in Wnt1-expressing tumors.
Our previous work provided evidence that Wnt/β-‐catenin signaling contributes to the BBB phenotype of the developing CNS through the transcriptional regulation of the tight junction protein claudin-‐3. Furthermore, the coverage of pericytes to brain vessels has been described to correlate with BBB integrity. In agreement with these publications, vessels of intracranial Wnt1-‐expressing GL261 tumors retained or regained barrier properties, indicated by a reduced leakage of the tracer Evans blue and endogenous mouse immunoglobulin G and increased junctional localiza2on of the tight junction proteins claudin-‐3, -‐5 and zonula occludens-‐1.
Overall, we detected sustained endothelial Wnt/β-‐catenin signaling to induce a quiescent and normalized tumor vascularization. Interestingly, the Notch signaling pathway has been shown to inhibit the angiogenic tip cell and to promote the quiescent stalk cell phenotype via its ligand Delta-like ligand 4 (Dll4) and the receptors Notch1 and 4. Mechanistically, my work demonstrated for the first time that overactivation of endothelial Wnt/β-‐catenin signaling reactivated expression of Dll4 in the tumor endothelium, which could be shown in vitro to increase Notch signaling and to favor a stalk cell-like gene signature. Furthermore, we uncovered the platelet-derived growth factor subunit B (pdgm) as a novel transcriptional target of Wnt/β-catenin signaling in ECs. Hence endothelial-‐derived PDGF-‐B is known to promote the recruitment of mural cells, the upregulation of this factor might explain the increased SMC/pericyte coverage observed in the tumor vasculature upon sustained endothelial Wnt/β-‐catenin signaling which additionally might promote a cycle of vascular normalization.
Taken together, my work reveals several vascular effects, being mediated by reinforced endothelial Wnt/β-‐catenin signaling during tumor angiogenesis. While a moderate level of canonical Wnt signaling, observed in vessels of human astrocytomas and murine control tumors, is considered to be associated with tumor angiogenesis, dominant activation of this pathway in ECs is shown to limit angiogenesis and to promote a quiescent and normalized tumor vasculature with increased barrier properties. Furthermore, my work discovers pdgm as a novel target of canonical Wnt signaling in ECs.
The work presented in this dissertation therefore not only uncovers the role of endothelial Wnt/β-‐catenin signaling in tumor angiogenesis but additionally reveals this pathway to be a novel modulator in pathological vessel development which might proof to be a valuable therapeutic target for anti-angiogenic and edema glioma therapy.
Im Zentralen Nervensystem (ZNS) kommunizieren neuronale Synapsen über eine Kombination von chemischen und elektrischen Signalen, die in ihrer Umgebung eine spezifische Komposition von Ionen benötigen. Um eine strenge Kontrolle des ZNS-Milieus zu gewährleisten, hat sich in Säugetieren eine endotheliale Blut-Hirn-Schranke (BHS) entwickelt. Die BHS limitiert den parazellulären Molekül Transport und wird von den Kapillargefässen des Gehirns gebildet, wobei die physische Barrier von den Tight Junctions (TJs) des vaskulären Endothels generiert wird. Das Gehirnendothel ist Teil einer neurovaskulären Einheit (NVE), zu der auch Perizyten (PZ), Astrozyten (AZ), Mikroglia und Interneurone zählen. Fehlkommunikation oder defekte zelluläre Komponenten in der NVE führen in der Regel zu Störungen in der BHS Funktion und können schwerwiegende neuronale Erkrankungen zur Folge haben.
Vor einigen Jahren haben wir und andere Forschungsgruppen herausgefunden, dass der Wnt/β-Catenin Signalweg essentiell für die Vaskularisierung des Gehirns während der Embryonalentwicklung ist und darüber hinaus auch eine bedeutende Rolle in der Induktion der BHS spielt. Des Weiteren konnte im Zebrafischmodell eine Aktivierung des kanonischen Wnt Signalweges auch im adulten Organismus nachgewiesen werden. Allerdings ist die Quelle der Wnt Wachstumsfaktoren bis dato unbekannt. Der Wnt Signalweg ist eine hoch konservierte und komplexe zelluläre Signalkaskade, die in allen mehrzelligen Organismen vorkommt. Wnt Wachstumsfaktoren sind sekretierte, hydrophobe Signalmoleküle, die sowohl über lange als auch kurze Strecken entweder den β-Catenin-abhängingen („kanonischen“) oder β-Catenin-unabhängingen („nicht-kanonischen“) Wnt Signalweg aktivieren können.
Da die meisten ZNS Erkrankungen mit einem Zusammenbruch der BHS-Funktion assoziiert sind, ist die Forschung bestrebt die Mechanismen, die der Entstehung und Aufrechterhaltung der BHS zugrunde liegen, zu ermitteln und zu verstehen. Das Ziel meiner Doktorarbeit war es herauszufinden, ob AZ Wnts produzieren und ob deren Wirkung auf das Gehirnendothel an der Aufrechterhaltung der BHS beteiligt ist. Zu diesem Zweck, habe ich ein in vitro BHS Kokultivierungs-Modellsystem etabliert das erstmalig ausschliesslich auf der Verwendung von murinen AZ und Gehirnendothelzellen basiert. Zu Beginn der Studie wurden sowohl primäre AZ als auch eine murine Gehirnendothel-zelllinie (MBE) bezüglich ihrer zell-spezifischen Eigenschaften charakterisiert. Dabei konnte belegt werden, dass sowohl die primären AZ als auch die MBE Zelllinie, aufgrund ihrer Proteinexpressionsprofile als repräsentative Vertreter ihres Zelltyps eingestuft werden können. Die darauffolgenden Untersuchungen konnten zeigen, dass primäre AZ über mehrere Passagen hinweg fast alle 19 Wnt Liganden auf mRNA Ebene exprimierten. Ferner konnte in primären Gehirnendothelzellen und zwei Gehirnendothelzelllinien die korrespondierenden Frizzled (FZD) Rezeptoren und low density lipoprotein receptor-related protein (LRP) Korezeptoren nachgewiesen werden. Dieser Befund legte Nahe, dass AZ und Gehirnendothelzellen die basalen Eigenschaften besitzen, um über den Wnt Signalweg miteinander zu kommunizieren. Die Stimulation von pMBEs mit Astrozyten konditioniertem Medium (AKM) induzierte die Hochregulation von Claudin-3 einem bekannten kanonischen Wnt Zielgens. Interessanterweise konnte diese Regulation teilweise durch die Zugabe von dickkopf 1 (Dkk1), einem Wnt/β-Catenin Antagonisten, inhibiert werden.
Um die physiologische Rolle der Wnt Liganden zu bestimmen, habe ich mir die Eigenschaft des universellen Sekretionsmechanismus der Wachstumsfaktoren, welcher von dem Transmembranprotein evenness interrupted (Evi) abhängig ist, zu Nutze gemacht. Die Verpaarung von Evifl/fl mit hGFAP-Cre Mäusen erlaubt die AZ-spezifische Deletion des Evi Proteins (Evi KO), was zur Folge hat, dass die Astrozyten der Nachkommen keine Wnt Wachstumsfaktoren sekretieren können.
In vitro führte der Verlust von Wnts in AKM zu einer teilweisen Delokalisierung von Junction Proteinen. Während die Kokultivierung mit Evi WT AZ einen straken Anstieg im TEER und reduzierte Permeabilitätsmesswerte induzierten, konnten diese pro-BHS Eigenschaften bei Evi KO AZ nicht beobachtet werden. Diese Ergebnisse zeigten deutlich, dass Wnts sekretiert von AZ den BHS Phenotyp positive beeinflussen, indem sie die Zell-Zell-Verbindung verstärken, was wiederum zu erhöhtem Zellwiderstand und reduzierter transzellulärer Permeabilität führt. Die Analyse des in vivo Phänotyps von Evi KO Mäusen ergab, dass mit fortschreitendem, postnatalem Alter makroskopisch erkennbare zerebrale Blutungen auftraten. Ausserdem konnte ich zeigen, dass eine Subpopulation von Blutgefässen Malformationen aufwies, die mit reduzierter Astrozytenendfuss-Assoziierung einhergingen.
Das Wissen um die Beteiligung des Wnt Signalweges an der Regulation der BHS auch im adulten Organismus kann in Zukunft von wichtiger Bedeutung sein, da es potentielle therapeutische Anwendungen ermöglicht.
The brain vascular system is composed of specialized endothelial cells, which regulate the movement of ions, molecules and cells from the blood lumen to the central nervous system (CNS). Endothelial cells in the brain form the blood-brain barrier (BBB) that is essential to maintain the brain homeostasis and protect the CNS from pathogens and toxins for a proper neurological function. Endothelium together with other cellular components such as pericytes, astrocytes and the basement membrane, forms the neurovascular unit (NVU), the structural unit of the BBB. Breakdown of the BBB occurs in various neurological disorders, leading to edema and neuronal damage. Therapeutic strategies focusing on factors that regulate the permeability of the BBB may help to improve neurological disorders and facilitate drug delivery to the brain.
Angiopoietins (Ang) are potential candidates for therapeutic targeting the BBB due to their role in regulating the vascular permeability in periphery. They are key growth factors that control angiogenesis and vessel maturation. Ang-1 and Ang-2 possess similar binding affinities to the Tie2 receptor tyrosine kinase, which is almost exclusively expressed on endothelial cells. Ang-1 is expressed in smooth muscle cells and pericytes, and binds in a paracrine manner to Tie2. This results in phosphorylation of the receptor and induction of downstream signaling pathways leading to vessel maturation via pericyte recruitment and blood vessel stabilization. Ang-2, on the other hand, is stored in Weibel Palade bodies in endothelial cells and is released upon inflammatory or angiogenic stimuli. Therefore, in mature, stabilized blood vessels, Ang-2 expression is low. Increased level of Ang-2 is only observed during development or in pathology such as ischemia, cancer and inflammation. When Ang-2 is released, it acts in an autocrine manner and interferes with Tie2 phosphorylation in a context-dependent way. Antagonizing the receptor results in de-stabilization of the vessels, often accompanied by reduced numbers of pericytes leading to myeloid cell infiltration. In conjunction with the vascular endothelial growth factor (VEGF), Ang-2 contributes to blood vessel sprouting, whereupon in absence of VEGF it promotes vessel regression. ...
